Mitigating Factual Hallucination in LRMs via Mixed-Mode Advantage Regularization (arxiv.org) 1 points by root-parent 1d ago ↗ HN
1 comment
[ 2.6 ms ] story [ 2.6 ms ] thread